584 Part IV / Perception
Figure 25–2 The double-step task illustrates how the brain
stabilizes images during saccades.
A.A subject starts by looking at a fixation point (FP) that disap-
pears, after which two saccade targets A and B appear and disap-
pear sequentially before the subject can make the saccade. The
first saccade (to target A) is simple. The retinal vector (FP→A) and
the saccade vectors are the same. After the first saccade, the
subject is looking at A. The retinal vector is A→B′, but the mon-
key must make a saccade whose vector is A→B. The brain must
adjust the retinal vector to compensate for the first saccade.
B.Timing. The upper records show when the targets appear
(colored bars). (Abbreviations: H, horizontal; V, vertical.)
A 任务几何 B 任务定时
200 毫秒
扫视到 A
扫视到 B
A
A
B
B
B′
注视点
注视点
水平
垂直
眼动
目标外观
problem. Every time a monkey makes a saccade, a
stimulus currently not in the receptive field of a neu-
ron in the lateral intraparietal area, and therefore inca-
pable of exciting the neuron, will excite the neuron if
the impending saccade will bring the stimulus into the
receptive field, even before the saccade occurs (Figure
25–3). Thus, a corollary discharge of the impending
saccade affects the visual responsiveness of the pari-
etal neuron.
This transient remapping of the receptive field
explains how subjects can perform the double-step task.
Consider the diagram in Figure 25–2A. The task begins
with the monkey directing gaze to the fixation point
(FP). After the monkey makes the first saccade, the reti-
nal vector A→B′ is no longer useful for making the A→B
saccade. However, the FP→A saccade remaps the activ-
ity of the cell describing the vector A→B, so it responds
to the target at the retinal location of B, which was not
in its receptive field when the monkey was looking at
FP. Remapping is found in a number of cortical and
subcortical areas, including lateral intraparietal area,
frontal eye field, medial intraparietal area, intermediate
layers of the superior colliculus, and prestriate areas V4,
V3a, and V2. As we shall see, remapping facilitates both
visual perception around the time of a saccade and the
accuracy of visually guided movement.
The first question this raises is: How does the brain
obtain the vector of the saccade that it feeds back to
the visual system? We know from decades of research
that the motor command for the vector is represented
in the superior colliculus on the roof of the midbrain
(Chapter 35). Each neuron in the superior colliculus
is tuned to saccades of a given vector, such that the
neurons collectively provide a map of the vectors of
all possible saccades. Inactivation of the superior col-
liculus affects the monkey’s ability to make saccades.
Electrical stimulation of the superior colliculus evokes
saccades of the vector described by the neurons at
the stimulation site. But this provides the vectors that
actually drive the eye, not the vectors that inform per-
ception about the vector of the saccade. How does the
vector information used to move the eye become avail-
able to brain processes that do not move the eye but do
require information about how it moved?
Since the vectors for moving the eye have been
identified in the superior colliculus, it is reasonable to
expect that this also might be the source of a corollary
discharge. Indeed, it is. The superior colliculus has both
descending pathways for generating the saccades and
ascending pathways to the cerebral cortex that could
carry the corollary discharge of the impending move-
ment (Figure 25–4). The pathways to the cortex pass
through the thalamus, as does all internal and almost
all external information reaching the cerebral cortex.
The motor signal in the thalamus is not necessar-
ily a corollary discharge; it could also be a movement
command that simply passes through the cerebral
cortex. That is not the case, however, because inactiva-
tion of this pathway in the thalamus does not alter the
amplitude and direction of saccades. It is not driving
saccades. It is more likely to be a corollary discharge.
After inactivation of the thalamic pathway, monkeys
cannot accurately perform the second saccade of the
double-step task. In addition, inactivation disrupts the
Kandel-Ch25_0582-0597.indd 584 19/01/21 4:03 PM